Components
Uptime, the long view
90 days · oldest on the left, today on the right · times in SAST
The last few hiccups, told plainly
Increased webhook delivery latency
Webhook deliveries to subscriber endpoints saw 8-12 second delays during peak hour. Root cause was an undersized worker pool after a recent traffic increase. We've doubled capacity and added an alert on queue depth.
- 15:08Resolved
Queue cleared and latency back to baseline. We're holding monitoring on for the rest of the day.
- 14:48Mitigating
Doubled the worker pool and started draining the backlog. Latency is recovering.
- 14:38Identified
Confirmed: queue worker pool is undersized for current load. Scaling up.
- 14:22Investigating
Investigating reports of delayed webhook deliveries.
Email & WhatsApp delivery delays
Brief 14-minute window during which transactional emails and WhatsApp messages queued without sending. Caused by a maintenance window on our messaging provider; messages flushed automatically once the window closed.
- 02:14Resolved
Provider window closed; queue flushed. No messages lost.
- 02:00Identified
Provider maintenance window in progress. Outbound messaging queued.
Scheduled maintenance
Webhook signing key rotation
We’re rotating the secret used to sign outbound webhooks. Existing signatures continue to verify for a 24-hour grace window. Integrations using the legacy v0 signing scheme should switch to v1 by 14 May.
Want to know when something changes?
We’ll email you the moment we open an incident, when we update it, and when it resolves. Unsubscribe with one click.